1 # This file is part of NIT ( http://www.nitlanguage.org ).
3 # This file is free software, which comes along with NIT. This software is
4 # distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY;
5 # without even the implied warranty of MERCHANTABILITY or FITNESS FOR A
6 # PARTICULAR PURPOSE. You can modify it is you want, provided this header
7 # is kept unaltered, and a notification of the changes is added.
8 # You are allowed to redistribute it and sell it, alone or is a part of
11 # Abstract class for manipulation of sequences of characters
17 intrude import collection
::array
25 # High-level abstraction for all text representations
29 redef type OTHER: Text
31 # Type of self (used for factorization of several methods, ex : substring_from, empty...)
34 # Gets a view on the chars of the Text object
36 # assert "hello".chars.to_a == ['h', 'e', 'l', 'l', 'o']
37 fun chars
: SequenceRead[Char] is abstract
39 # Gets a view on the bytes of the Text object
41 # assert "hello".bytes.to_a == [104u8, 101u8, 108u8, 108u8, 111u8]
42 fun bytes
: SequenceRead[Byte] is abstract
44 # Number of characters contained in self.
46 # assert "12345".length == 5
47 # assert "".length == 0
48 # assert "あいうえお".length == 5
49 fun length
: Int is abstract
51 # Number of bytes in `self`
53 # assert "12345".bytelen == 5
54 # assert "あいうえお".bytelen == 15
55 fun bytelen
: Int is abstract
59 # assert "abcd".substring(1, 2) == "bc"
60 # assert "abcd".substring(-1, 2) == "a"
61 # assert "abcd".substring(1, 0) == ""
62 # assert "abcd".substring(2, 5) == "cd"
63 # assert "あいうえお".substring(1,3) == "いうえ"
65 # A `from` index < 0 will be replaced by 0.
66 # Unless a `count` value is > 0 at the same time.
67 # In this case, `from += count` and `count -= from`.
68 fun substring
(from
: Int, count
: Int): SELFTYPE is abstract
70 # Iterates on the substrings of self if any
71 private fun substrings
: Iterator[FlatText] is abstract
73 # Is the current Text empty (== "")
76 # assert not "foo".is_empty
77 fun is_empty
: Bool do return self.length
== 0
79 # Returns an empty Text of the right type
81 # This method is used internally to get the right
82 # implementation of an empty string.
83 protected fun empty
: SELFTYPE is abstract
85 # Gets the first char of the Text
87 # DEPRECATED : Use self.chars.first instead
88 fun first
: Char do return self.chars
[0]
90 # Access a character at `index` in the string.
92 # assert "abcd"[2] == 'c'
94 # DEPRECATED : Use self.chars.[] instead
95 fun [](index
: Int): Char do return self.chars
[index
]
97 # Gets the index of the first occurence of 'c'
99 # Returns -1 if not found
101 # DEPRECATED : Use self.chars.index_of instead
102 fun index_of
(c
: Char): Int
104 return index_of_from
(c
, 0)
107 # Gets the last char of self
109 # DEPRECATED : Use self.chars.last instead
110 fun last
: Char do return self.chars
[length-1
]
112 # Gets the index of the first occurence of ´c´ starting from ´pos´
114 # Returns -1 if not found
116 # DEPRECATED : Use self.chars.index_of_from instead
117 fun index_of_from
(c
: Char, pos
: Int): Int
119 var iter
= self.chars
.iterator_from
(pos
)
121 if iter
.item
== c
then return iter
.index
127 # Gets the last index of char ´c´
129 # Returns -1 if not found
131 # DEPRECATED : Use self.chars.last_index_of instead
132 fun last_index_of
(c
: Char): Int
134 return last_index_of_from
(c
, length
- 1)
137 # Return a null terminated char *
138 fun to_cstring
: NativeString is abstract
140 # The index of the last occurrence of an element starting from pos (in reverse order).
142 # var s = "/etc/bin/test/test.nit"
143 # assert s.last_index_of_from('/', s.length-1) == 13
144 # assert s.last_index_of_from('/', 12) == 8
146 # Returns -1 if not found
148 # DEPRECATED : Use self.chars.last_index_of_from instead
149 fun last_index_of_from
(item
: Char, pos
: Int): Int do return chars
.last_index_of_from
(item
, pos
)
151 # Gets an iterator on the chars of self
153 # DEPRECATED : Use self.chars.iterator instead
154 fun iterator
: Iterator[Char]
156 return self.chars
.iterator
160 # Gets an Array containing the chars of self
162 # DEPRECATED : Use self.chars.to_a instead
163 fun to_a
: Array[Char] do return chars
.to_a
165 # Create a substring from `self` beginning at the `from` position
167 # assert "abcd".substring_from(1) == "bcd"
168 # assert "abcd".substring_from(-1) == "abcd"
169 # assert "abcd".substring_from(2) == "cd"
171 # As with substring, a `from` index < 0 will be replaced by 0
172 fun substring_from
(from
: Int): SELFTYPE
174 if from
>= self.length
then return empty
175 if from
< 0 then from
= 0
176 return substring
(from
, length
- from
)
179 # Does self have a substring `str` starting from position `pos`?
181 # assert "abcd".has_substring("bc",1) == true
182 # assert "abcd".has_substring("bc",2) == false
184 # Returns true iff all characters of `str` are presents
185 # at the expected index in `self.`
186 # The first character of `str` being at `pos`, the second
187 # character being at `pos+1` and so on...
189 # This means that all characters of `str` need to be inside `self`.
191 # assert "abcd".has_substring("xab", -1) == false
192 # assert "abcd".has_substring("cdx", 2) == false
194 # And that the empty string is always a valid substring.
196 # assert "abcd".has_substring("", 2) == true
197 # assert "abcd".has_substring("", 200) == true
198 fun has_substring
(str
: String, pos
: Int): Bool
200 if str
.is_empty
then return true
201 if pos
< 0 or pos
+ str
.length
> length
then return false
202 var myiter
= self.chars
.iterator_from
(pos
)
203 var itsiter
= str
.chars
.iterator
204 while myiter
.is_ok
and itsiter
.is_ok
do
205 if myiter
.item
!= itsiter
.item
then return false
209 if itsiter
.is_ok
then return false
213 # Is this string prefixed by `prefix`?
215 # assert "abcd".has_prefix("ab") == true
216 # assert "abcbc".has_prefix("bc") == false
217 # assert "ab".has_prefix("abcd") == false
218 fun has_prefix
(prefix
: String): Bool do return has_substring
(prefix
,0)
220 # Is this string suffixed by `suffix`?
222 # assert "abcd".has_suffix("abc") == false
223 # assert "abcd".has_suffix("bcd") == true
224 fun has_suffix
(suffix
: String): Bool do return has_substring
(suffix
, length
- suffix
.length
)
226 # Returns `self` as the corresponding integer
228 # assert "123".to_i == 123
229 # assert "-1".to_i == -1
230 # assert "0x64".to_i == 100
231 # assert "0b1100_0011".to_i== 195
232 # assert "--12".to_i == 12
234 # REQUIRE: `self`.`is_int`
235 fun to_i
: Int is abstract
237 # If `self` contains a float, return the corresponding float
239 # assert "123".to_f == 123.0
240 # assert "-1".to_f == -1.0
241 # assert "-1.2e-3".to_f == -0.0012
245 return to_s
.to_cstring
.atof
248 # If `self` contains only digits and alpha <= 'f', return the corresponding integer.
250 # assert "ff".to_hex == 255
251 fun to_hex
(pos
, ln
: nullable Int): Int do
253 if pos
== null then pos
= 0
254 if ln
== null then ln
= length
- pos
256 for i
in [pos
.. max
[ do
258 res
+= self[i
].from_hex
263 # If `self` contains only digits <= '7', return the corresponding integer.
265 # assert "714".to_oct == 460
266 fun to_oct
: Int do return a_to
(8)
268 # If `self` contains only '0' et '1', return the corresponding integer.
270 # assert "101101".to_bin == 45
271 fun to_bin
: Int do return a_to
(2)
273 # If `self` contains only digits '0' .. '9', return the corresponding integer.
275 # assert "108".to_dec == 108
276 fun to_dec
: Int do return a_to
(10)
278 # If `self` contains only digits and letters, return the corresponding integer in a given base
280 # assert "120".a_to(3) == 15
281 fun a_to
(base
: Int) : Int
286 for j
in [0..length
[ do
308 # Is this string in a valid numeric format compatible with `to_f`?
310 # assert "123".is_numeric == true
311 # assert "1.2".is_numeric == true
312 # assert "-1.2".is_numeric == true
313 # assert "1,2".is_numeric == true
314 # assert "-1.23e-2".is_numeric == true
315 # assert "1..2".is_numeric == false
316 # assert "".is_numeric == false
319 var has_point_or_comma
= false
321 for i
in [0..length
[ do
323 if not c
.is_numeric
then
324 if (c
== '.' or c
== ',') and not has_point_or_comma
then
325 has_point_or_comma
= true
326 else if c
== 'e' and e_index
== -1 and i
> 0 and i
< length
- 1 and chars
[i-1
] != '-' then
328 else if c
== '-' and i
== e_index
+ 1 and i
< length
- 1 then
337 # Returns `true` if the string contains only Hex chars
339 # assert "048bf".is_hex == true
340 # assert "ABCDEF".is_hex == true
341 # assert "0G".is_hex == false
344 for i
in [0..length
[ do
346 if not (c
>= 'a' and c
<= 'f') and
347 not (c
>= 'A' and c
<= 'F') and
348 not (c
>= '0' and c
<= '9') then return false
353 # Returns `true` if the string contains only Binary digits
355 # assert "1101100".is_bin == true
356 # assert "1101020".is_bin == false
358 for i
in chars
do if i
!= '0' and i
!= '1' then return false
362 # Returns `true` if the string contains only Octal digits
364 # assert "213453".is_oct == true
365 # assert "781".is_oct == false
367 for i
in chars
do if i
< '0' or i
> '7' then return false
371 # Returns `true` if the string contains only Decimal digits
373 # assert "10839".is_dec == true
374 # assert "164F".is_dec == false
376 for i
in chars
do if i
< '0' or i
> '9' then return false
380 # Are all letters in `self` upper-case ?
382 # assert "HELLO WORLD".is_upper == true
383 # assert "%$&%!".is_upper == true
384 # assert "hello world".is_upper == false
385 # assert "Hello World".is_upper == false
388 for i
in [0..length
[ do
390 if char
.is_lower
then return false
395 # Are all letters in `self` lower-case ?
397 # assert "hello world".is_lower == true
398 # assert "%$&%!".is_lower == true
399 # assert "Hello World".is_lower == false
402 for i
in [0..length
[ do
404 if char
.is_upper
then return false
409 # Removes the whitespaces at the beginning of self
411 # assert " \n\thello \n\t".l_trim == "hello \n\t"
413 # `Char::is_whitespace` determines what is a whitespace.
416 var iter
= self.chars
.iterator
418 if not iter
.item
.is_whitespace
then break
421 if iter
.index
== length
then return self.empty
422 return self.substring_from
(iter
.index
)
425 # Removes the whitespaces at the end of self
427 # assert " \n\thello \n\t".r_trim == " \n\thello"
429 # `Char::is_whitespace` determines what is a whitespace.
432 var iter
= self.chars
.reverse_iterator
434 if not iter
.item
.is_whitespace
then break
437 if iter
.index
< 0 then return self.empty
438 return self.substring
(0, iter
.index
+ 1)
441 # Trims trailing and preceding white spaces
443 # assert " Hello World ! ".trim == "Hello World !"
444 # assert "\na\nb\tc\t".trim == "a\nb\tc"
446 # `Char::is_whitespace` determines what is a whitespace.
447 fun trim
: SELFTYPE do return (self.l_trim
).r_trim
449 # Is the string non-empty but only made of whitespaces?
451 # assert " \n\t ".is_whitespace == true
452 # assert " hello ".is_whitespace == false
453 # assert "".is_whitespace == false
455 # `Char::is_whitespace` determines what is a whitespace.
456 fun is_whitespace
: Bool
458 if is_empty
then return false
459 for c
in self.chars
do
460 if not c
.is_whitespace
then return false
465 # Returns `self` removed from its last line terminator (if any).
467 # assert "Hello\n".chomp == "Hello"
468 # assert "Hello".chomp == "Hello"
470 # assert "\n".chomp == ""
471 # assert "".chomp == ""
473 # Line terminators are `"\n"`, `"\r\n"` and `"\r"`.
474 # A single line terminator, the last one, is removed.
476 # assert "\r\n".chomp == ""
477 # assert "\r\n\n".chomp == "\r\n"
478 # assert "\r\n\r\n".chomp == "\r\n"
479 # assert "\r\n\r".chomp == "\r\n"
481 # Note: unlike with most IO methods like `Reader::read_line`,
482 # a single `\r` is considered here to be a line terminator and will be removed.
486 if len
== 0 then return self
487 var l
= self.chars
.last
489 return substring
(0, len-1
)
490 else if l
!= '\n' then
492 else if len
> 1 and self.chars
[len-2
] == '\r' then
493 return substring
(0, len-2
)
495 return substring
(0, len-1
)
499 # Justify a self in a space of `length`
501 # `left` is the space ratio on the left side.
502 # * 0.0 for left-justified (no space at the left)
503 # * 1.0 for right-justified (all spaces at the left)
504 # * 0.5 for centered (half the spaces at the left)
508 # assert "hello".justify(10, 0.0) == "hello "
509 # assert "hello".justify(10, 1.0) == " hello"
510 # assert "hello".justify(10, 0.5) == " hello "
512 # If `length` is not enough, `self` is returned as is.
514 # assert "hello".justify(2, 0.0) == "hello"
516 # REQUIRE: `left >= 0.0 and left <= 1.0`
517 # ENSURE: `self.length <= length implies result.length == length`
518 # ENSURE: `self.length >= length implies result == self`
519 fun justify
(length
: Int, left
: Float): String
521 var diff
= length
- self.length
522 if diff
<= 0 then return to_s
523 assert left
>= 0.0 and left
<= 1.0
524 var before
= (diff
.to_f
* left
).to_i
525 return " " * before
+ self + " " * (diff-before
)
528 # Mangle a string to be a unique string only made of alphanumeric characters and underscores.
530 # This method is injective (two different inputs never produce the same
531 # output) and the returned string always respect the following rules:
533 # * Contains only US-ASCII letters, digits and underscores.
534 # * Never starts with a digit.
535 # * Never ends with an underscore.
536 # * Never contains two contiguous underscores.
538 # assert "42_is/The answer!".to_cmangle == "_52d2_is_47dThe_32danswer_33d"
539 # assert "__".to_cmangle == "_95d_95d"
540 # assert "__d".to_cmangle == "_95d_d"
541 # assert "_d_".to_cmangle == "_d_95d"
542 # assert "_42".to_cmangle == "_95d42"
543 # assert "foo".to_cmangle == "foo"
544 # assert "".to_cmangle == ""
545 fun to_cmangle
: String
547 if is_empty
then return ""
549 var underscore
= false
553 if c
>= '0' and c
<= '9' then
555 res
.append
(c
.code_point
.to_s
)
559 for i
in [start
..length
[ do
561 if (c
>= 'a' and c
<= 'z') or (c
>='A' and c
<= 'Z') then
567 res
.append
('_'.code_point
.to_s
)
570 if c
>= '0' and c
<= '9' then
573 else if c
== '_' then
578 res
.append
(c
.code_point
.to_s
)
584 res
.append
('_'.code_point
.to_s
)
590 # Escape " \ ' and non printable characters using the rules of literal C strings and characters
592 # assert "abAB12<>&".escape_to_c == "abAB12<>&"
593 # assert "\n\"'\\".escape_to_c == "\\n\\\"\\'\\\\"
595 # Most non-printable characters (bellow ASCII 32) are escaped to an octal form `\nnn`.
596 # Three digits are always used to avoid following digits to be interpreted as an element
597 # of the octal sequence.
599 # assert "{0.code_point}{1.code_point}{8.code_point}{31.code_point}{32.code_point}".escape_to_c == "\\000\\001\\010\\037 "
601 # The exceptions are the common `\t` and `\n`.
602 fun escape_to_c
: String
605 for i
in [0..length
[ do
609 else if c
== '\t' then
611 else if c
== '"' then
613 else if c == '\'' then
615 else if c == '\\
' then
617 else if c.code_point < 32 then
619 var oct = c.code_point.to_base(8, false)
620 # Force 3 octal digits since it is the
621 # maximum allowed in the C specification
622 if oct.length == 1 then
625 else if oct.length == 2 then
636 # Escape additionnal characters
637 # The result might no be legal in C but be used in other languages
639 # assert "ab|\{\}".escape_more_to_c("|\{\}") == "ab\\|\\\{\\\}"
640 fun escape_more_to_c(chars: String): String
643 for c in escape_to_c.chars do
644 if chars.chars.has(c) then
652 # Escape to C plus braces
654 # assert "\n\"'\\\
{\}".escape_to_nit == "\\n\\\
"\\'\\\\\\\{\\\}"
655 fun escape_to_nit
: String do return escape_more_to_c
("\{\}")
657 # Escape to POSIX Shell (sh).
659 # Abort if the text contains a null byte.
661 # assert "\n\"'\\\{\}0".escape_to_sh == "'\n\"'\\''\\\{\}0'"
662 fun escape_to_sh
: String do
665 for i in [0..length[ do
670 assert without_null_byte
: c
!= '\0'
678 # Escape to include in a Makefile
680 # Unfortunately, some characters are not escapable in Makefile.
681 # These characters are `;`, `|`, `\`, and the non-printable ones.
682 # They will be rendered as `"?{hex}"`.
683 fun escape_to_mk: String do
685 for i in [0..length[ do
689 else if c == ':' or c == ' ' or c == '#' then
692 else if c
.code_point
< 32 or c
== ';' or c
== '|' or c
== '\\' or c
== '=' then
693 b
.append
("?{c.code_point.to_base(16, false)}")
701 # Return a string where Nit escape sequences are transformed.
704 # assert s.length == 2
705 # var u = s.unescape_nit
706 # assert u.length == 1
707 # assert u.chars[0].code_point == 10 # (the ASCII value of the "new line" character)
708 fun unescape_nit
: String
710 var res
= new Buffer.with_cap
(self.length
)
711 var was_slash
= false
712 for i
in [0..length
[ do
714 if not was_slash
then
725 else if c
== 'r' then
727 else if c
== 't' then
729 else if c
== '0' then
738 # Returns `self` with all characters escaped with their UTF-16 representation
740 # assert "Aèあ𐏓".escape_to_utf16 == "\\u0041\\u00e8\\u3042\\ud800\\udfd3"
741 fun escape_to_utf16
: String do
743 for i
in chars
do buf
.append i
.escape_to_utf16
747 # Returns the Unicode char escaped by `self`
749 # assert "\\u0041".from_utf16_escape == 'A'
750 # assert "\\ud800\\udfd3".from_utf16_escape == '𐏓'
751 # assert "\\u00e8".from_utf16_escape == 'è'
752 # assert "\\u3042".from_utf16_escape == 'あ'
753 fun from_utf16_escape
(pos
, ln
: nullable Int): Char do
754 if pos
== null then pos
= 0
755 if ln
== null then ln
= length
- pos
756 if ln
< 6 then return 0xFFFD.code_point
757 var cp
= from_utf16_digit
(pos
+ 2)
758 if cp
< 0xD800 then return cp
.code_point
759 if cp
> 0xDFFF then return cp
.code_point
760 if cp
> 0xDBFF then return 0xFFFD.code_point
761 if ln
== 6 then return 0xFFFD.code_point
762 if ln
< 12 then return 0xFFFD.code_point
764 cp
+= from_utf16_digit
(pos
+ 8)
765 var cplo
= cp
& 0xFFFF
766 if cplo
< 0xDC00 then return 0xFFFD.code_point
767 if cplo
> 0xDFFF then return 0xFFFD.code_point
768 return cp
.from_utf16_surr
.code_point
771 # Returns a UTF-16 escape value
773 # var s = "\\ud800\\udfd3"
774 # assert s.from_utf16_digit(2) == 0xD800
775 # assert s.from_utf16_digit(8) == 0xDFD3
776 fun from_utf16_digit
(pos
: nullable Int): Int do
777 if pos
== null then pos
= 0
778 return to_hex
(pos
, 4)
781 # Encode `self` to percent (or URL) encoding
783 # assert "aBc09-._~".to_percent_encoding == "aBc09-._~"
784 # assert "%()< >".to_percent_encoding == "%25%28%29%3c%20%3e"
785 # assert ".com/post?e=asdf&f=123".to_percent_encoding == ".com%2fpost%3fe%3dasdf%26f%3d123"
786 # assert "éあいう".to_percent_encoding == "%c3%a9%e3%81%82%e3%81%84%e3%81%86"
787 fun to_percent_encoding
: String
791 for i
in [0..length
[ do
793 if (c
>= '0' and c
<= '9') or
794 (c
>= 'a' and c
<= 'z') or
795 (c
>= 'A' and c
<= 'Z') or
796 c
== '-' or c
== '.' or
801 var bytes
= c
.to_s
.bytes
802 for b
in bytes
do buf
.append
"%{b.to_i.to_hex}"
809 # Decode `self` from percent (or URL) encoding to a clear string
811 # Replace invalid use of '%' with '?'.
813 # assert "aBc09-._~".from_percent_encoding == "aBc09-._~"
814 # assert "%25%28%29%3c%20%3e".from_percent_encoding == "%()< >"
815 # assert ".com%2fpost%3fe%3dasdf%26f%3d123".from_percent_encoding == ".com/post?e=asdf&f=123"
816 # assert "%25%28%29%3C%20%3E".from_percent_encoding == "%()< >"
817 # assert "incomplete %".from_percent_encoding == "incomplete ?"
818 # assert "invalid % usage".from_percent_encoding == "invalid ? usage"
819 # assert "%c3%a9%e3%81%82%e3%81%84%e3%81%86".from_percent_encoding == "éあいう"
820 fun from_percent_encoding
: String
823 var has_percent
= false
831 # If no transformation is needed, return self as a string
832 if not has_percent
then return to_s
834 var buf
= new NativeString(len
)
840 if i
+ 2 >= length
then
841 # What follows % has been cut off
845 var hex_s
= substring
(i
, 2)
847 var hex_i
= hex_s
.to_hex
851 # What follows a % is not Hex
856 else buf
[l
] = c
.ascii
862 return buf
.to_s_unsafe
(l
)
865 # Escape the characters `<`, `>`, `&`, `"`, `'` and `/` as HTML/XML entity references.
867 # assert "a&b-<>\"x\"/'".html_escape == "a&b-<>"x"/'"
869 # SEE: <https://www.owasp.org/index.php/XSS_%28Cross_Site_Scripting%29_Prevention_Cheat_Sheet#RULE_.231_-_HTML_Escape_Before_Inserting_Untrusted_Data_into_HTML_Element_Content>
870 fun html_escape
: String
874 for i
in [0..length
[ do
878 else if c
== '<' then
880 else if c
== '>' then
882 else if c
== '"' then
884 else if c
== '\'' then
886 else if c == '/' then
895 # Two pieces of text are equals if thez have the same characters in the same order.
897 # assert "hello" == "hello"
898 # assert "hello" != "HELLO"
899 # assert "hello" == "hel"+"lo"
901 # Things that are not Text are not equal.
904 # assert "9" != ['9']
907 # assert "9".chars.first == '9' # equality of Char
908 # assert "9".chars == ['9'] # equality of Sequence
909 # assert "9".to_i == 9 # equality of Int
912 if o == null then return false
913 if not o isa Text then return false
914 if self.is_same_instance(o) then return true
915 if self.length != o.length then return false
916 return self.chars == o.chars
919 # Lexicographical comparaison
921 # assert "abc" < "xy"
922 # assert "ABC" < "abc"
925 var self_chars = self.chars.iterator
926 var other_chars = other.chars.iterator
928 while self_chars.is_ok and other_chars.is_ok do
929 if self_chars.item < other_chars.item then return true
930 if self_chars.item > other_chars.item then return false
935 if self_chars.is_ok then
942 # Escape string used in labels for graphviz
944 # assert ">><<".escape_to_dot == "\\>\\>\\<\\<"
945 fun escape_to_dot: String
947 return escape_more_to_c("|\{\}<>")
950 private var hash_cache: nullable Int = null
954 if hash_cache == null then
955 # djb2 hash algorithm
958 for i in [0..length[ do
960 h = (h << 5) + h + char.code_point
965 return hash_cache.as(not null)
968 # Gives the formatted string back as a Nit string with `args` in place
970 # assert "This %1 is a %2.".format("String", "formatted String") == "This String is a formatted String."
971 # assert "\\%1 This string".format("String") == "\\%1 This string"
972 fun format(args: Object...): String do
973 var s = new Array[Text]
977 # Skip escaped characters
978 if self[i] == '\\
' then
981 else if self[i] == '%' then
985 while i < length and self[i].is_numeric do
990 var ciph_len = fmt_end - ciph_st + 1
992 var arg_index = substring(ciph_st, ciph_len).to_i - 1
993 if arg_index >= args.length then continue
995 s.push substring(curr_st, fmt_st - curr_st)
996 s.push args[arg_index].to_s
1001 s.push substring(curr_st, length - curr_st)
1005 # Return the Levenshtein distance between two strings
1008 # assert "abcd".levenshtein_distance("abcd") == 0
1009 # assert "".levenshtein_distance("abcd") == 4
1010 # assert "abcd".levenshtein_distance("") == 4
1011 # assert "abcd".levenshtein_distance("xyz") == 4
1012 # assert "abcd".levenshtein_distance("xbdy") == 3
1014 fun levenshtein_distance(other: String): Int
1016 var slen = self.length
1017 var olen = other.length
1020 if slen == 0 then return olen
1021 if olen == 0 then return slen
1022 if self == other then return 0
1024 # previous row of distances
1025 var v0 = new Array[Int].with_capacity(olen+1)
1027 # current row of distances
1028 var v1 = new Array[Int].with_capacity(olen+1)
1030 for j in [0..olen] do
1031 # prefix insert cost
1035 for i in [0..slen[ do
1037 # prefix delete cost
1040 for j in [0..olen[ do
1042 var cost1 = v1[j] + 1
1044 var cost2 = v0[j + 1] + 1
1045 # same char cost (+0)
1048 if self[i] != other[j] then cost3 += 1
1050 v1[j+1] = cost1.min(cost2).min(cost3)
1054 # * v1 become v0 in the next iteration
1055 # * old v0 is reused as the new v1
1064 # Copies `n` bytes from `self` at `src_offset` into `dest` starting at `dest_offset`
1066 # Basically a high-level synonym of NativeString::copy_to
1068 # REQUIRE: `n` must be large enough to contain `len` bytes
1070 # var ns = new NativeString(8)
1071 # "Text is String".copy_to_native(ns, 8, 2, 0)
1072 # assert ns.to_s_unsafe(8) == "xt is St"
1074 fun copy_to_native(dest: NativeString, n, src_offset, dest_offset: Int) do
1075 var mypos = src_offset
1076 var itspos = dest_offset
1078 dest[itspos] = self.bytes[mypos]
1087 # All kinds of array-based text representations.
1088 abstract class FlatText
1091 # Underlying C-String (`char*`)
1093 # Warning : Might be void in some subclasses, be sure to check
1094 # if set before using it.
1095 var items: NativeString is noinit
1097 # Returns a char* starting at position `first_byte`
1099 # WARNING: If you choose to use this service, be careful of the following.
1101 # Strings and NativeString are *ideally* always allocated through a Garbage Collector.
1102 # Since the GC tracks the use of the pointer for the beginning of the char*, it may be
1103 # deallocated at any moment, rendering the pointer returned by this function invalid.
1104 # Any access to freed memory may very likely cause undefined behaviour or a crash.
1105 # (Failure to do so will most certainly result in long and painful debugging hours)
1107 # The only safe use of this pointer is if it is ephemeral (e.g. read in a C function
1108 # then immediately return).
1110 # As always, do not modify the content of the String in C code, if this is what you want
1111 # copy locally the char* as Nit Strings are immutable.
1112 fun fast_cstring: NativeString is abstract
1114 redef var length = 0
1116 redef var bytelen = 0
1127 redef fun copy_to_native(dest, n, src_offset, dest_offset) do
1128 items.copy_to(dest, n, src_offset, dest_offset)
1132 # Abstract class for the SequenceRead compatible
1133 # views on the chars of any Text
1134 private abstract class StringCharView
1135 super SequenceRead[Char]
1139 var target: SELFTYPE
1141 redef fun is_empty do return target.is_empty
1143 redef fun length do return target.length
1145 redef fun iterator: IndexedIterator[Char] do return self.iterator_from(0)
1147 redef fun reverse_iterator do return self.reverse_iterator_from(self.length - 1)
1150 # Abstract class for the SequenceRead compatible
1151 # views on the bytes of any Text
1152 private abstract class StringByteView
1153 super SequenceRead[Byte]
1157 var target: SELFTYPE
1159 redef fun is_empty do return target.is_empty
1161 redef fun length do return target.bytelen
1163 redef fun iterator do return self.iterator_from(0)
1165 redef fun reverse_iterator do return self.reverse_iterator_from(target.bytelen - 1)
1168 # Immutable sequence of characters.
1170 # String objects may be created using literals.
1172 # assert "Hello World!" isa String
1173 abstract class String
1176 redef type SELFTYPE: String is fixed
1178 redef fun to_s do return self
1180 # Concatenates `o` to `self`
1182 # assert "hello" + "world" == "helloworld"
1183 # assert "" + "hello" + "" == "hello"
1184 fun +(o: Text): SELFTYPE is abstract
1186 # Concatenates self `i` times
1188 # assert "abc" * 4 == "abcabcabcabc"
1189 # assert "abc" * 1 == "abc"
1190 # assert "abc" * 0 == ""
1191 fun *(i: Int): SELFTYPE is abstract
1193 # Insert `s` at `pos`.
1195 # assert "helloworld".insert_at(" ", 5) == "hello world"
1196 fun insert_at(s: String, pos: Int): SELFTYPE is abstract
1198 redef fun substrings is abstract
1200 # Returns a reversed version of self
1202 # assert "hello".reversed == "olleh"
1203 # assert "bob".reversed == "bob"
1204 # assert "".reversed == ""
1205 fun reversed: SELFTYPE is abstract
1207 # A upper case version of `self`
1209 # assert "Hello World!".to_upper == "HELLO WORLD!"
1210 fun to_upper: SELFTYPE is abstract
1212 # A lower case version of `self`
1214 # assert "Hello World!".to_lower == "hello world!"
1215 fun to_lower : SELFTYPE is abstract
1217 # Takes a camel case `self` and converts it to snake case
1219 # assert "randomMethodId".to_snake_case == "random_method_id"
1221 # The rules are the following:
1223 # An uppercase is always converted to a lowercase
1225 # assert "HELLO_WORLD".to_snake_case == "hello_world"
1227 # An uppercase that follows a lowercase is prefixed with an underscore
1229 # assert "HelloTheWORLD".to_snake_case == "hello_the_world"
1231 # An uppercase that follows an uppercase and is followed by a lowercase, is prefixed with an underscore
1233 # assert "HelloTHEWorld".to_snake_case == "hello_the_world"
1235 # All other characters are kept as is; `self` does not need to be a proper CamelCased string.
1237 # assert "=-_H3ll0Th3W0rld_-=".to_snake_case == "=-_h3ll0th3w0rld_-="
1238 fun to_snake_case: SELFTYPE
1240 if self.is_lower then return self
1242 var new_str = new Buffer.with_cap(self.length)
1243 var prev_is_lower = false
1244 var prev_is_upper = false
1246 for i in [0..length[ do
1248 if char.is_lower then
1250 prev_is_lower = true
1251 prev_is_upper = false
1252 else if char.is_upper then
1253 if prev_is_lower then
1255 else if prev_is_upper and i+1 < length and chars[i+1].is_lower then
1258 new_str.add(char.to_lower)
1259 prev_is_lower = false
1260 prev_is_upper = true
1263 prev_is_lower = false
1264 prev_is_upper = false
1271 # Takes a snake case `self` and converts it to camel case
1273 # assert "random_method_id".to_camel_case == "randomMethodId"
1275 # If the identifier is prefixed by an underscore, the underscore is ignored
1277 # assert "_private_field".to_camel_case == "_privateField"
1279 # If `self` is upper, it is returned unchanged
1281 # assert "RANDOM_ID".to_camel_case == "RANDOM_ID"
1283 # If there are several consecutive underscores, they are considered as a single one
1285 # assert "random__method_id".to_camel_case == "randomMethodId"
1286 fun to_camel_case: SELFTYPE
1288 if self.is_upper then return self
1290 var new_str = new Buffer
1291 var is_first_char = true
1292 var follows_us = false
1294 for i in [0..length[ do
1296 if is_first_char then
1298 is_first_char = false
1299 else if char == '_
' then
1301 else if follows_us then
1302 new_str.add(char.to_upper)
1312 # Returns a capitalized `self`
1314 # Letters that follow a letter are lowercased
1315 # Letters that follow a non-letter are upcased.
1317 # SEE : `Char::is_letter` for the definition of letter.
1319 # assert "jAVASCRIPT".capitalized == "Javascript"
1320 # assert "i am root".capitalized == "I Am Root"
1321 # assert "ab_c -ab0c ab\nc".capitalized == "Ab_C -Ab0C Ab\nC"
1322 fun capitalized: SELFTYPE do
1323 if length == 0 then return self
1325 var buf = new Buffer.with_cap(length)
1327 var curr = chars[0].to_upper
1331 for i in [1 .. length[ do
1334 if prev.is_letter then
1335 buf[i] = curr.to_lower
1337 buf[i] = curr.to_upper
1345 # A mutable sequence of characters.
1346 abstract class Buffer
1349 # Returns an arbitrary subclass of `Buffer` with default parameters
1352 # Returns an instance of a subclass of `Buffer` with `i` base capacity
1353 new with_cap(i: Int) is abstract
1355 redef type SELFTYPE: Buffer is fixed
1357 # Specific implementations MUST set this to `true` in order to invalidate caches
1358 protected var is_dirty = true
1360 # Copy-On-Write flag
1362 # If the `Buffer` was to_s'd
, the next in-place altering
1363 # operation will cause the current `Buffer` to be re-allocated.
1365 # The flag will then be set at `false`.
1366 protected var written
= false
1368 # Modifies the char contained at pos `index`
1370 # DEPRECATED : Use self.chars.[]= instead
1371 fun []=(index
: Int, item
: Char) is abstract
1373 # Adds a char `c` at the end of self
1375 # DEPRECATED : Use self.chars.add instead
1376 fun add
(c
: Char) is abstract
1380 # var b = new Buffer
1382 # assert not b.is_empty
1385 fun clear
is abstract
1387 # Enlarges the subsequent array containing the chars of self
1388 fun enlarge
(cap
: Int) is abstract
1390 # Adds the content of text `s` at the end of self
1392 # var b = new Buffer
1395 # assert b == "helloworld"
1396 fun append
(s
: Text) is abstract
1398 # `self` is appended in such a way that `self` is repeated `r` times
1400 # var b = new Buffer
1403 # assert b == "hellohellohello"
1404 fun times
(r
: Int) is abstract
1406 # Reverses itself in-place
1408 # var b = new Buffer
1411 # assert b == "olleh"
1412 fun reverse
is abstract
1414 # Changes each lower-case char in `self` by its upper-case variant
1416 # var b = new Buffer
1417 # b.append("Hello World!")
1419 # assert b == "HELLO WORLD!"
1420 fun upper
is abstract
1422 # Changes each upper-case char in `self` by its lower-case variant
1424 # var b = new Buffer
1425 # b.append("Hello World!")
1427 # assert b == "hello world!"
1428 fun lower
is abstract
1430 # Capitalizes each word in `self`
1432 # Letters that follow a letter are lowercased
1433 # Letters that follow a non-letter are upcased.
1435 # SEE: `Char::is_letter` for the definition of a letter.
1437 # var b = new FlatBuffer.from("jAVAsCriPt")
1439 # assert b == "Javascript"
1440 # b = new FlatBuffer.from("i am root")
1442 # assert b == "I Am Root"
1443 # b = new FlatBuffer.from("ab_c -ab0c ab\nc")
1445 # assert b == "Ab_C -Ab0C Ab\nC"
1447 if length
== 0 then return
1448 var c
= self[0].to_upper
1451 for i
in [1 .. length
[ do
1454 if prev
.is_letter
then
1455 self[i
] = c
.to_lower
1457 self[i
] = c
.to_upper
1464 if is_dirty
then hash_cache
= null
1468 # In Buffers, the internal sequence of character is mutable
1469 # Thus, `chars` can be used to modify the buffer.
1470 redef fun chars
: Sequence[Char] is abstract
1473 # View for chars on Buffer objects, extends Sequence
1474 # for mutation operations
1475 private abstract class BufferCharView
1476 super StringCharView
1477 super Sequence[Char]
1479 redef type SELFTYPE: Buffer
1483 # View for bytes on Buffer objects, extends Sequence
1484 # for mutation operations
1485 private abstract class BufferByteView
1486 super StringByteView
1488 redef type SELFTYPE: Buffer
1492 # User readable representation of `self`.
1493 fun to_s
: String do return inspect
1495 # The class name of the object in NativeString format.
1496 private fun native_class_name
: NativeString is intern
1498 # The class name of the object.
1500 # assert 5.class_name == "Int"
1501 fun class_name
: String do return native_class_name
.to_s
1503 # Developer readable representation of `self`.
1504 # Usually, it uses the form "<CLASSNAME:#OBJECTID bla bla bla>"
1507 return "<{inspect_head}>"
1510 # Return "CLASSNAME:#OBJECTID".
1511 # This function is mainly used with the redefinition of the inspect method
1512 protected fun inspect_head
: String
1514 return "{class_name}:#{object_id.to_hex}"
1519 # assert true.to_s == "true"
1520 # assert false.to_s == "false"
1532 # C function to calculate the length of the `NativeString` to receive `self`
1533 private fun byte_to_s_len
: Int `{
1534 return snprintf(NULL, 0, "0x%02x", self);
1537 # C function to convert an nit Int to a NativeString (char*)
1538 private fun native_byte_to_s
(nstr
: NativeString, strlen
: Int) `{
1539 snprintf(nstr, strlen, "0x%02x", self);
1542 # Displayable byte in its hexadecimal form (0x..)
1544 # assert 1.to_b.to_s == "0x01"
1545 # assert (-123).to_b.to_s == "0x85"
1547 var nslen
= byte_to_s_len
1548 var ns
= new NativeString(nslen
+ 1)
1550 native_byte_to_s
(ns
, nslen
+ 1)
1551 return ns
.to_s_unsafe
(nslen
)
1557 # Wrapper of strerror C function
1558 private fun strerror_ext
: NativeString `{ return strerror(self); `}
1560 # Returns a string describing error number
1561 fun strerror: String do return strerror_ext.to_s
1563 # Fill `s
` with the digits in base `base
` of `self` (and with the '-' sign if 'signed' and negative).
1564 # assume < to_c max const of char
1565 private fun fill_buffer(s: Buffer, base: Int, signed: Bool)
1572 else if self == 0 then
1579 var pos = digit_count(base) - 1
1580 while pos >= 0 and n > 0 do
1581 s.chars[pos] = (n % base).to_c
1587 # C function to calculate the length of the `NativeString` to receive `self`
1588 private fun int_to_s_len: Int `{
1589 return snprintf
(NULL, 0, "%ld", self);
1592 # C function to convert an nit Int to a NativeString (char*)
1593 private fun native_int_to_s(nstr: NativeString, strlen: Int) `{
1594 snprintf
(nstr
, strlen
, "%ld", self);
1597 # return displayable int in base base and signed
1598 fun to_base(base: Int, signed: Bool): String is abstract
1600 # return displayable int in hexadecimal
1602 # assert 1.to_hex == "1"
1603 # assert (-255).to_hex == "-ff"
1604 fun to_hex: String do return to_base(16,false)
1608 # Pretty representation of `self`, with decimals as needed from 1 to a maximum of 3
1610 # assert 12.34.to_s == "12.34"
1611 # assert (-0120.030).to_s == "-120.03"
1613 # see `to_precision
` for a custom precision.
1615 var str = to_precision( 3 )
1616 if is_inf != 0 or is_nan then return str
1617 var len = str.length
1618 for i in [0..len-1] do
1620 var c = str.chars[j]
1623 else if c == '.' then
1624 return str.substring( 0, j+2 )
1626 return str.substring( 0, j+1 )
1632 # `String` representation of `self` with the given number of `decimals
`
1634 # assert 12.345.to_precision(0) == "12"
1635 # assert 12.345.to_precision(3) == "12.345"
1636 # assert (-12.345).to_precision(3) == "-12.345"
1637 # assert (-0.123).to_precision(3) == "-0.123"
1638 # assert 0.999.to_precision(2) == "1.00"
1639 # assert 0.999.to_precision(4) == "0.9990"
1640 fun to_precision(decimals: Int): String
1642 if is_nan then return "nan"
1644 var isinf = self.is_inf
1647 else if isinf == -1 then
1651 if decimals == 0 then return self.to_i.to_s
1653 for i in [0..decimals[ do f = f * 10.0
1660 if i == 0 then return "0." + "0"*decimals
1662 # Prepare both parts of the float, before and after the "."
1667 if sl > decimals then
1668 # Has something before the "."
1669 p1 = s.substring(0, sl-decimals)
1670 p2 = s.substring(sl-decimals, decimals)
1673 p2 = "0"*(decimals-sl) + s
1676 if i < 0 then p1 = "-" + p1
1678 return p1 + "." + p2
1684 # Returns a sequence with the UTF-8 bytes of `self`
1686 # assert 'a'.bytes == [0x61u8]
1687 # assert 'ま'.bytes == [0xE3u8, 0x81u8, 0xBEu8]
1688 fun bytes: SequenceRead[Byte] do return to_s.bytes
1690 # Is `self` an UTF-16 surrogate pair ?
1691 fun is_surrogate: Bool do
1693 return cp >= 0xD800 and cp <= 0xDFFF
1696 # Length of `self` in a UTF-8 String
1697 private fun u8char_len: Int do
1698 var c = self.code_point
1699 if c < 0x80 then return 1
1700 if c <= 0x7FF then return 2
1701 if c <= 0xFFFF then return 3
1702 if c <= 0x10FFFF then return 4
1703 # Bad character format
1707 # assert 'x'.to_s == "x"
1710 var ns = new NativeString(ln + 1)
1712 return ns.to_s_unsafe(ln)
1715 # Returns `self` escaped to UTF-16
1717 # i.e. Represents `self`.`code_point
` using UTF-16 codets escaped
1720 # assert 'A'.escape_to_utf16 == "\\u0041"
1721 # assert 'è'.escape_to_utf16 == "\\u00e8"
1722 # assert 'あ'.escape_to_utf16 == "\\u3042"
1723 # assert '𐏓'.escape_to_utf16 == "\\ud800\\udfd3"
1724 fun escape_to_utf16: String do
1727 if cp < 0xD800 or (cp >= 0xE000 and cp <= 0xFFFF) then
1728 buf = new Buffer.with_cap(6)
1729 buf.append("\\u0000")
1732 for i in hx.chars.reverse_iterator do
1737 buf = new Buffer.with_cap(12)
1738 buf.append("\\u0000\\u0000")
1739 var lo = (((cp - 0x10000) & 0x3FF) + 0xDC00).to_hex
1740 var hi = ((((cp - 0x10000) & 0xFFC00) >> 10) + 0xD800).to_hex
1755 private fun u8char_tos(r: NativeString, len: Int) `{
1762 r
[0] = 0xC0 | ((self & 0x7C0) >> 6);
1763 r
[1] = 0x80 | (self & 0x3F);
1766 r
[0] = 0xE0 | ((self & 0xF000) >> 12);
1767 r
[1] = 0x80 | ((self & 0xFC0) >> 6);
1768 r
[2] = 0x80 | (self & 0x3F);
1771 r
[0] = 0xF0 | ((self & 0x1C0000) >> 18);
1772 r
[1] = 0x80 | ((self & 0x3F000) >> 12);
1773 r
[2] = 0x80 | ((self & 0xFC0) >> 6);
1774 r
[3] = 0x80 | (self & 0x3F);
1779 # Returns true if the char is a numerical digit
1781 # assert '0'.is_numeric
1782 # assert '9'.is_numeric
1783 # assert not 'a'.is_numeric
1784 # assert not '?'.is_numeric
1786 # FIXME: Works on ASCII-range only
1787 fun is_numeric: Bool
1789 return self >= '0' and self <= '9'
1792 # Returns true if the char is an alpha digit
1794 # assert 'a'.is_alpha
1795 # assert 'Z'.is_alpha
1796 # assert not '0'.is_alpha
1797 # assert not '?'.is_alpha
1799 # FIXME: Works on ASCII-range only
1802 return (self >= 'a' and self <= 'z') or (self >= 'A' and self <= 'Z')
1805 # Is `self` an hexadecimal digit ?
1807 # assert 'A'.is_hexdigit
1808 # assert not 'G'.is_hexdigit
1809 # assert 'a'.is_hexdigit
1810 # assert not 'g'.is_hexdigit
1811 # assert '5'.is_hexdigit
1812 fun is_hexdigit: Bool do return (self >= '0' and self <= '9') or (self >= 'A' and self <= 'F') or
1813 (self >= 'a' and self <= 'f')
1815 # Returns true if the char is an alpha or a numeric digit
1817 # assert 'a'.is_alphanumeric
1818 # assert 'Z'.is_alphanumeric
1819 # assert '0'.is_alphanumeric
1820 # assert '9'.is_alphanumeric
1821 # assert not '?'.is_alphanumeric
1823 # FIXME: Works on ASCII-range only
1824 fun is_alphanumeric: Bool
1826 return self.is_numeric or self.is_alpha
1829 # Returns `self` to its int value
1831 # REQUIRE: `is_hexdigit
`
1832 fun from_hex: Int do
1833 if self >= '0' and self <= '9' then return code_point - 0x30
1834 if self >= 'A' and self <= 'F' then return code_point - 0x37
1835 if self >= 'a' and self <= 'f' then return code_point - 0x57
1836 # Happens if self is not a hexdigit
1837 assert self.is_hexdigit
1838 # To make flow analysis happy
1843 redef class Collection[E]
1844 # String representation of the content of the collection.
1846 # The standard representation is the list of elements separated with commas.
1849 # assert [1,2,3].to_s == "[1,2,3]"
1850 # assert [1..3].to_s == "[1,2,3]"
1851 # assert (new Array[Int]).to_s == "[]" # empty collection
1854 # Subclasses may return a more specific string representation.
1857 return "[" + join(",") + "]"
1860 # Concatenate elements without separators
1863 # assert [1,2,3].plain_to_s == "123"
1864 # assert [11..13].plain_to_s == "111213"
1865 # assert (new Array[Int]).plain_to_s == "" # empty collection
1867 fun plain_to_s: String
1870 for e in self do if e != null then s.append(e.to_s)
1874 # Concatenate and separate each elements with `separator
`.
1876 # Only concatenate if `separator
== null`.
1878 # assert [1, 2, 3].join(":") == "1:2:3"
1879 # assert [1..3].join(":") == "1:2:3"
1880 # assert [1..3].join == "123"
1881 fun join(separator: nullable Text): String
1883 if is_empty then return ""
1885 var s = new Buffer # Result
1890 if e != null then s.append(e.to_s)
1892 # Concat other items
1895 if separator != null then s.append(separator)
1897 if e != null then s.append(e.to_s)
1904 redef class Map[K,V]
1905 # Concatenate couples of key value.
1906 # Key and value are separated by `couple_sep
`.
1907 # Couples are separated by `sep
`.
1909 # var m = new HashMap[Int, String]
1912 # assert m.join("; ", "=") == "1=one; 10=ten"
1913 fun join(sep, couple_sep: String): String is abstract
1917 private var args_cache: nullable Sequence[String] = null
1919 # The arguments of the program as given by the OS
1920 fun program_args: Sequence[String]
1922 if _args_cache == null then init_args
1923 return _args_cache.as(not null)
1926 # The name of the program as given by the OS
1927 fun program_name: String
1929 return native_argv(0).to_s
1932 # Initialize `program_args
` with the contents of `native_argc
` and `native_argv
`.
1933 private fun init_args
1935 var argc = native_argc
1936 var args = new Array[String].with_capacity(0)
1939 args[i-1] = native_argv(i).to_s
1945 # First argument of the main C function.
1946 private fun native_argc: Int is intern
1948 # Second argument of the main C function.
1949 private fun native_argv(i: Int): NativeString is intern
1952 # Comparator that efficienlty use `to_s
` to compare things
1954 # The comparaison call `to_s
` on object and use the result to order things.
1956 # var a = [1, 2, 3, 10, 20]
1957 # (new CachedAlphaComparator).sort(a)
1958 # assert a == [1, 10, 2, 20, 3]
1960 # Internally the result of `to_s
` is cached in a HashMap to counter
1961 # uneficient implementation of `to_s
`.
1963 # Note: it caching is not usefull, see `alpha_comparator
`
1964 class CachedAlphaComparator
1966 redef type COMPARED: Object
1968 private var cache = new HashMap[Object, String]
1970 private fun do_to_s(a: Object): String do
1971 if cache.has_key(a) then return cache[a]
1977 redef fun compare(a, b) do
1978 return do_to_s(a) <=> do_to_s(b)
1982 # see `alpha_comparator
`
1983 private class AlphaComparator
1985 redef fun compare(a, b) do return a.to_s <=> b.to_s
1988 # Stateless comparator that naively use `to_s
` to compare things.
1990 # Note: the result of `to_s
` is not cached, thus can be invoked a lot
1991 # on a single instace. See `CachedAlphaComparator` as an alternative.
1993 # var a = [1, 2, 3, 10, 20]
1994 # alpha_comparator.sort(a)
1995 # assert a == [1, 10, 2, 20, 3]
1996 fun alpha_comparator: Comparator do return once new AlphaComparator
1998 # The arguments of the program as given by the OS
1999 fun args: Sequence[String]
2001 return sys.program_args
2004 redef class NativeString
2005 # Returns `self` as a new String.
2006 fun to_s_with_copy: String is abstract
2008 # Returns `self` as a String of `length
`.
2009 fun to_s_with_length(length: Int): String is abstract
2011 # Returns a new instance of `String` with self as `_items
`
2013 # /!\: Does not clean the items for compliance with UTF-8,
2014 # Use only if you know what you are doing
2015 fun to_s_unsafe(len: nullable Int): String is abstract
2017 # Returns `self` as a String with `bytelen
` and `length
` set
2019 # SEE: `abstract_text
::Text` for more infos on the difference
2020 # between `Text::bytelen
` and `Text::length
`
2021 fun to_s_full(bytelen, unilen: Int): String is abstract
2024 redef class NativeArray[E]
2025 # Join all the elements using `to_s
`
2027 # REQUIRE: `self isa NativeArray[String]`
2028 # REQUIRE: all elements are initialized
2029 fun native_to_s: String is abstract